World of Warcraft

1 . 2 . 3 . 4 . 5 . 6 . 7 . 8
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 0. Hit Rating Theorycraft   10-May-2007 18:53:57 PDT
quote reply
Lately, i've seen many posts in which ppl ask things about HIT so i thought to make a post in order to clear up the things about hit. In the next lines i'll try to analyze what exactly the hit rating does. Keep in mind that assumptions and numbers refer to fights against raid bosses since they are the most important thing in raids and refer only to the current situation since things may change a bit after the patch due to glancing blow reduction.

NOTES:
1) Raid bosses from what we know so far seem to be 73 lvl so their defence is 365.
After several tests, it has been proven that the current miss chance against raid bosses is most likely to be 9% if you use 2H or 28% if you dual wield. Here is where the weapon skill comes. According to the latest tests, 2 formulas have been created to determine the role of weapon skill in missing. The formulas are the following ones:

* If the difference between weapon skill and mob's defence is equal or lower than 10 then the formula for calculating the miss chance against a mob is: 5% + 0.1%*X where X is the difference between weapon skill and defence and therefore takes values 1-10. If you are dw, the 5% is replaced by 24%

* If the difference between weapon skill and mob's defence is greater than 10 then the formula to calculate the miss chance against a mob is: 7% + 0.4%*(Mob's defence - weapon skill - 10). If you are dw, 7% is replaced by 26%.

Why 7% in the second formula and not 5%?. A small example:

You fight against a boss. The boss has 365 defence and you have 350 weapon skill. The base miss chance is 5%. However, you get 1% miss penalty which rises it to 6% because the difference is greater than 10. Now, since you have 15 points of difference, the first 10 points increase your miss chance by 0.1% per each point giving you a total of 1% miss chance so the 6% becomes 7%. The rest 5 points increase your miss chance by 0.4% per point.

You add 2 points of weapon skill. Now the difference is 365-352=13 which is obviously greater than 10. Again, you have the 5% + 1% (10*0.1 for the first 10 points) + 1% miss penalty + 3*0.4%.

Summarizing these 2 above examples we see how the 2nd formula results.

**SWEET POINT**: Looking at the above formulas, we see that there is a sweet point. This is when you reduce the difference between weapon skill and defence to 10 ( assuming it's 11 or more). At this point, you get the 1% hit jump. In a few words, if you have 354 weapon skill, your miss chance against a boss is 7.4%. If you add 1 more point of weapon skill rising it up to 355 your new miss chance will be 6%. This makes the racial bonuses very important in terms of missing. An Orc will wielding axe will have 3% less miss chance than a tauren wielding the same axe due to the Orc racial.

Also, weapon skill lvls are dropped and not rounded. That said, if you have +15 mace skill rating this equals to 3.8% weapon skill but it's dropped down to 3% and not rounded up to 4%.

***As of 2.3, Weapon Skill no longer exists and it's replaced by Weapon Expertise which actually reduces the chance that your attacks will be dodged or parried.***

2) Attacking a mob from behind can remove only parry and block from the table since direction doesn't affect dodge. Also note that there is a 0.5 yard penalty in which direction doesn't count.

3) Attacks divide in 2 categories: Auto ( white) attacks and Special attacks ( Mortal Strike, Sinister Strike etc). Auto attacks use the 1 roll theory while special attacks use the 2 roll theory. However, BASE miss chance seems to be the same for auto and special attacks. Special attacks don't suffer the DW penalty.

4) 1 Roll theory: The basic idea on which the 1 roll system relies is that every melee (auto) attack made by players against mobs can have only one of the next possible results:
Miss
Dodge
Parry
Glancing Blow
Block
Critical Hit
Ordinary Hit

The following tables are laid out in descending order of the precedence of one attack result over another. In simple words, the entries at the top of the table take precedence over the entries below them. The precedence order is given from Blizzard so we can safely consider it accurate.

2 Roll theory: The basic idea of the 2 roll theory is that every melee (special) attack made by players against mobs can have only one of the next possible results:
Miss
Dodge
Parry
Block
Ordinary Hit

After the roll for the attack is performed the server performs another internal action/roll which seems something like this:

"If value rolled is block then
roll for blocked hit or blocked crit
If value rolled is hit then
roll for ordinary hit or critical hit".

The next images prove that special attacks are more likely to be based on a 2 roll theory:
http://www.layoutsource.net/uploads/images/2007-05-11/3rWAsQqlof.jpg
http://www.layoutsource.net/uploads/images/2007-05-11/PqHUlIw1YM.jpg

These were some important notes that we must keep in mind in order to continue. Now, what does exactly hit does? The answer is not what most ppl think and it's a bit complicated. Stacking hit doesn't directly increase your chance to hit!
HIT DECREASES YOUR CHANCE TO MISS. So, how HIT interacts with auto attacks??

Let's take an example: Let's assume that you are a dual wielder and you fight against a raid boss, your hit is 0% and your crit is 50%. Also assume that your weapon skill is maxed out for your lvl, 350@70 lvl and you're fighting the boss from the back. Since you are a dual wielder the game will regenerate the next table :
Miss: 28%
Dodge: 5.6%
Glancing Blow: 25%
Critical Hit: 41.4%
Ordinary Hit: 0%

This table shows that 8.6% of your crit was wasted. Now, Increase your hit by 10% and decrease your crit by 5% for example. Your new hit is 10% and overall crit is 45%. The new table will look like this one:
Miss: 18%
Dodge: 5.6%
Glancing Blow: 25%
Critical Hit: 45%
Ordinary Hit: 6.4%

As you notice, even if you lowered your overall crit rate for to add some hit, your actual crit rate was increased and no crit was wasted. In other words, HIT also increases your crit capacity. This example though, was refering only in auto attacks. However, the crit cap was much lower before the reduction of glancing blows and the difference was way much more obvious than now.

SPECIAL ATTACKS: Special attacks use a 2 roll system so the crit cap appears to be 100% making Hit less important for them and no important at all after covering the base miss chance which is 9%( no matter if you are a 2hander or dw). Now, let's say you mouseover your crit chance and it says you have 25% crit chance and you also don't have any %hit on you. Your special attacks crit chance against a raid boss (from the back) will be 25% of your hit which makes it to be 25% of 85.4%=21.35% since you don't have any %hit. Increase your hit by 10%. With +10% hit you have covered the base miss chance which is 9% so the rest 1% will be wasted. Your new special attacks crit chance will be 25% out of 94.4%=23.6%. So basically, HIT INCREASES YOUR CRIT CAP FOR AUTO ATTACKS AND UNTIL YOU COVER THE BASE MISS CHANCE IT WILL INCREASE YOUR CRIT CHANCE OF SPECIAL ATTACKS TOO.

Hit appears to be the most important stat till you reach a point where no crit will be wasted. After this point it's not so clear whether hit crit or ap benefits more. More hit will always increase your dps till you reach the hit cap. Crit and ap will also boost your dmg. So, when it comes to choose, it's always about the ammounts that you have to compare. 60 AP will probably offer more dps than 5 hit rating and 5 hit rating will probably offer more dps than 10 ap.

I hope i covered some questions about hit and it's role, however feel free to provide more infos if you think that i missed something or if something from the things mentioned above is wrong.

EDIT: I changed the first image since it was pvp related and it could lead to wrong results. I went in Ulda, tried it myself and captured 2 screenshots. Now example is pve based as it should be. More infos about the test on post #25, page 2. Also, i changed the miss chances according to the new tests that were done and posted in the EJ forums. Glancing blow chance is changed since tests and spreadsheets indicate a new rate of 25%. Link for Miss chance calculations and weapon skill assumptions:
http://elitistjerks.com/f31/t11885-rogue_warrior_weapon_skill_adjustment_discussion/p16/

[ Post edited by Cryinfreeman ]

70
View All Posts by This User Toggle Ignore / Unignore This User
  • 1. Re: Hit Rating Theorycraft   10-May-2007 19:35:15 PDT
quote reply
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 2. Re: Hit Rating Theorycraft   11-May-2007 01:17:20 PDT
quote reply
Thanks for clearing this up :)

It's nice that you do make the reservation that no crit must be wasted in order for hit to be that good. As in, once you reach a point where you actually do get regular hits, +hit isn't priority one stat anymore.

And I didn't know about the specials 2-roll system, I always learn new things :)

I get better all the time, fighting nerfs with gear upgrades.
Warr: http://tinyurl.com/2yydbc - Lock: http://tinyurl.com/28epf8
64
View All Posts by This User Toggle Ignore / Unignore This User
  • 3. Re: Hit Rating Theorycraft   11-May-2007 01:33:17 PDT
quote reply
Thanks for the interesting post.

The 2-roll system for special attacks is only for player attacks right? Or can special attacks from a mob also crit despite being blocked?
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 4. Re: Hit Rating Theorycraft   11-May-2007 01:53:10 PDT
quote reply

Q u o t e:
However, miss chance seems to be the same for auto and special attacks


incorrect. instant attacks and on next swing attacks do not incur the DW +miss penalty.
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 5. Re: Hit Rating Theorycraft   11-May-2007 02:09:59 PDT
quote reply
The 2 roll system must be refering only to attacks made by players. From what i know so far boss special attacks can never be a crit.

Yeah Michiko, I didn't post it right. The miss chance was refering to the base miss which is 6.5% no matter if you dw or 2h. Some ppl say that the base miss chance between special and auto attacks is different. This has not been proved yet so i kept for my examples the case where the base miss chance is the same.

Post edited :D
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 6. Re: Hit Rating Theorycraft   11-May-2007 02:32:08 PDT
quote reply
Using a 2 hander the base miss rate is higher than you report here. Even with +7% hit you will see a number of misses in a reasonably long boss fight.
I'v slowly increased my hit to run tests over several days and found i had to up my +hit to around 8.10% before i stop seeing misses.
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 7. Re: Hit Rating Theorycraft   11-May-2007 02:44:58 PDT
quote reply
As a rogue I can confirm that this is indeed the way duel wield works. You need about +25% to hit, before you stop missing autoattacks.

Also important, the -19% to hit penalty on auto attacks is on BOTH your weapons, not just the offhand, as many for some reason assume.

Heroic Trashmob Asscleaves you for 11723
You die
You give up
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 8. Re: Hit Rating Theorycraft   11-May-2007 03:07:59 PDT
quote reply
Very nice post, very helpfull as im one of the ones asking questions about hit rating vs crit and ap.

However, may I ask you for some conclusions? Im not that good at math, so if someone could help would be great. At what point, will youre hit be "capped" and should there be aimed for higher crit / ap? Many say 200 hit rating + talents, or is it slightly higher seeing pve raid bosses are higher level?

Thanks in advance should you take the time to elaborate!
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 9. Re: Hit Rating Theorycraft   11-May-2007 04:22:03 PDT
quote reply

Q u o t e:

SPECIAL ATTACKS: Special attacks use a 2 roll system

proof / blue post ?
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 10. Re: Hit Rating Theorycraft   11-May-2007 05:12:16 PDT
quote reply
This is a nice post and I like the job you put on it, but I do have questions and corrections. Please answer them.

A long time ago there was a big discussion about hit% and what you really needed back at lvl 60 vs bosses. Nothing has changed since then ;).


Q u o t e:
Mobs have their chance to be missed increased by 0.1% for every point that their defence exceeds player's weapon skill


I want to see the math behind it because mobs defences work just like players defence.
0.04% miss for every point of defense above the weapon skill of the opponent.

The base miss is 5%, adding 0,2% for a mobs level above you for 5,6%.

Now what was found was that from 6% there was still a chance to miss and it was not until you hit 8-9% that miss was almost completely removed from your special attacks (and one weapon wield).
So we had somekind of diminished returns that no one could explain, math was correct yet didn't add up.
-Note back then we had only full % not rating, so you we could only go 8% then 9%. Nothing inbetween.

Then the Dawn of AQ40 and the Twin Emperors came.
A rogue came with calculation on weapon skill on glancing blows, which was taken further by a few warriors. Hours upon ours of combat logs versus the Twin Emperors and a result was found for +weapon skill.

You needed 10 weapon skill to reach 0.6% hit!
Now with testing of almost 10 weapon skill and 5% hit the final result that that was the magic maker for eliminating miss from special attacks.
-Weapon skill effect on glancing blows was removed, not it's other effects on hit%.

-Your take on the miss% on higher level mobs is flawed ;).

- You made a mistake in your two roll theory list where you omitted "Critical hit" from the list.

- I would like to know where you got the 0.5 yard penalty and exactly what kind of penalty needs to be defined.

- Also define where you got your ideas on the two roll theory. I've seen this mention recently, yet I've seen nothing defnining proof or extrapolated data.

Edit:
Too bad that the whole "Glancing Blow" section on www.evilempireguild.org was removed since that is where all the data was held. Ciderhelm really should restore that section for it's information (even though it's outdated). But that is nostalgy on my part :p.

[ Post edited by Daylanor ]


( \_/) If I could give you only one advice, wear sunscreen would be it.
( ._.) /
((*)*)

For all things....Fury is here to save the day!
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 11. Re: Hit Rating Theorycraft   11-May-2007 06:05:12 PDT
quote reply

Q u o t e:

- Also define where you got your ideas on the two roll theory. I've seen this mention recently, yet I've seen nothing defnining proof or extrapolated data.



The link provided in the OP, shows a combat log of a blocked heroic strike crit. This suggest that either a) its a 2-roll system or b) pvp works differently.

Other than that, i havent done enough research and testing on hit to be able to say anything coclusively.

[ Post edited by Zuluhed ]


The road to hell is paved with blood elf paladins!
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 12. Re: Hit Rating Theorycraft   11-May-2007 06:13:31 PDT
quote reply
The only time where crits cannot be blocked is when mobs are attacking a player. Player crits vs mobs can be blocked and player crits vs players can also be blocked.

Sweeping Strikes Guide - http://surl.se/eli (update postponed due to 2.1.0)

SWORDS, NOT WORDS
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 13. Re: Hit Rating Theorycraft   11-May-2007 06:15:53 PDT
quote reply

Q u o t e:


The link provided in the OP, shows a combat log of a blocked heroic strike crit. This suggest that either a) its a 2-roll system or b) pvp works differently.

Other than that, i havent done enough research and testing on hit to be able to say anything coclusively.


PvP has always had slightly different rules.
When you speak of PvE you only use pve encounters for the data accumilation and work on that data only ;).

( \_/) If I could give you only one advice, wear sunscreen would be it.
( ._.) /
((*)*)

For all things....Fury is here to save the day!
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 14. Re: Hit Rating Theorycraft   11-May-2007 06:18:08 PDT
quote reply

Q u o t e:
The only time where crits cannot be blocked is when mobs are attacking a player. Player crits vs mobs can be blocked and player crits vs players can also be blocked.


What Meadows is saying is that if you block a mobs attack it can not be a critical attack.

( \_/) If I could give you only one advice, wear sunscreen would be it.
( ._.) /
((*)*)

For all things....Fury is here to save the day!
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 16. Re: Hit Rating Theorycraft   11-May-2007 08:17:40 PDT
quote reply
I'll try to explain some things mentioned above.

1 Roll theory: Under this theory, an attack can only have one possible result: miss/dodge/parry/glancing blow/block/crit/hit. So, if you make an attack against a mob can lead to ONLY AND ONLY ONE of the above results.

Take another example: 5/9 05:17:48.796 Your Riposte crits Olaf for 394. (22 blocked)
5/9 05:18:03.078 Your Sinister Strike crits Olaf for 549. (22 blocked)

These 2 actions were taken from a rogue's combat log while fighting Olaf in Ulda. Olaf has something special. He has 100% chance to block. Now, under the 1 roll theory, we shouldn't really see any crits. Since Olaf has a 100% block chance against incoming attacks, this means that crit and normal hits should be removed from the table. However, we still notice crits on the table which makes special attacks more likely to be based on a 2 roll theory. REMEMBER that it's NOT YET confirmed that special attacks use a 2 roll theory but tests prove that they do.


Q u o t e:
Using a 2 hander the base miss rate is higher than you report here. Even with +7% hit you will see a number of misses in a reasonably long boss fight.
I'v slowly increased my hit to run tests over several days and found i had to up my +hit to around 8.10% before i stop seeing misses.


Well, there is no way that miss chance for 2h is about 8.1% since no math formulas can prove it. It's also said that no matter what's your %hit, you will still see misses, their chance will be 1% or less but you'll still see misses ( not proven yet too). The only way to have valid results will be a test with a number of at least 5k swings against a 73 lvl mob. Also, there is a theory claiming that raid bosses have 440 def instead of 365 which makes you need 8.6% to theorytically remove misses of the table. If mob defence and player defence behave exactly in the same way it means that you need 8.6% hit for not to theoritically miss and it also leads to the conclusion that their chance to parry/block/dodge an attack would be 8.6% each lowering your crit cap by 15%. Wowwiki also says: Each point of +Defense skill adds 0.04% to the chance to be Missed, to Parry, to Dodge, and to Block for players. Mobs receive 0.1% per point of defense. So, either wiki is wrong or mob defence doesn't behave like the player defence. Nothing is sure yet, all these are tests and assumptions. The best would be a clear response from blizzard.


Q u o t e:
However, may I ask you for some conclusions? Im not that good at math, so if someone could help would be great. At what point, will youre hit be "capped" and should there be aimed for higher crit / ap? Many say 200 hit rating + talents, or is it slightly higher seeing pve raid bosses are higher level?


Raid bosses are 3 lvls higher than your lvl. This is confirmed by Blizzard according to wowwiki. This means that if you are 60 lvl and you fight a boss in MC, the boss for you will be 63 lvl. If you are 70 lvl, the boss for you will be 73 lvl. Your hit will be capped at the point that you will never see a miss which seems to be 25.5% hit if you dw or 6.5% (?) if you are a 2h. Of course, after a point, stacking hit will lead you to lower your crit and ap by far which is not worthy. What i wanted to say in my first post is that hit will never be useless till you reach your hit cap. After a point, it will be less important than crit or ap. We can't say for sure how much hit is needed. It seems like a 220 hit+ talents would be enough.


Q u o t e:
I want to see the math behind it because mobs defences work just like players defence.
0.04% miss for every point of defense above the weapon skill of the opponent.

The base miss is 5%, adding 0,2% for a mobs level above you for 5,6%.


As i stated before, boss defence either behaves like player defence does, or mobs have a different way in which defence works. If it behaves like the player's does, then 5.6% could be reasonable. Still, dw with 24.64% hit notice misses after long fights. This can be explained in 2 ways: First way is that no matter your hit chance you will still notice misses. The second way is that mobs gain 0.1% miss for each point that defence exceeds weapon skill so their actual chance to be missed is more than 24.6% which is expected. These are theories and assumptions and as i said before, the best would be an answer from Blizzard. Till we get it, it will all remain assumptions.

There is a 0.5 yard penalty due to lag issues etc. This means that if you hit a mob from behind and you are within this range, direction won't count and it will be like you hit boss from the front.


Q u o t e:
- You made a mistake in your two roll theory list where you omitted "Critical hit" from the list.


I didn't. The basic idea of the 2 roll system is that an attack can result in a miss/dodge/parry/block/ordinary hit. After this, the server performs another internal roll as i described in my first post. Let's implement the table with crit too. It will become:
Miss
Dodge
Parry
Block
Critical Hit
Ordinary Hit

You perform an attack and let's say that the result was a blocked crit. Under your theory, when you perform an attack, you actually have 2 different percentages of crit. The first percentage is the one of the critical hit which is on the table and the other one is an "internal" percentage which lead us to the blocked crit ( roll for first result, if block then roll for blocked hit or blocked crit). So, in simple words, you say that when you make an attack, your chance to crit will be either 30% or 25%( numbers are just an example) which i think it's pretty much impossible. Correct me though if i misunderstood your theory.

The idea of the 2 roll system came in my mind when i noticed that my special attacks crit rate was lower than the auto attack one. I used Lunar Crescent ( 3.7 weapon speed) to make sure that out of 1000 (i made a test with 1k swings) swings the number of special and auto attacks would be approximately the same. After the test i made, the number of my special attacks was a bit bigger than the auto attacks but still, the crit rate was lower. I've discussed it with many ppl and it seems like that the 2 roll theory is working as i stated. There is no clear proof though. It seems like the special attacks use a 2 roll system. It seems that the 2 roll system probably works as i described. Yet, an answer from Blizzard would make things more clear.

I hope i answered your questions. Still, if you think that i might be wrong somewhere feel free to add with your comments and corrections.
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 17. Re: Hit Rating Theorycraft   11-May-2007 09:29:23 PDT
quote reply

Q u o t e:
I'll try to explain some things mentioned above.

1 Roll theory: Under this theory, an attack can only have one possible result: miss/dodge/parry/glancing blow/block/crit/hit. So, if you make an attack against a mob can lead to ONLY AND ONLY ONE of the above results.

Take another example: 5/9 05:17:48.796 Your Riposte crits Olaf for 394. (22 blocked)
5/9 05:18:03.078 Your Sinister Strike crits Olaf for 549. (22 blocked)

bla-bla-bla


PvE and PvP use DIFFERENT roll mechanics.
in pvp you CAN see blocked crit.
screen of such miracle in pve plz.
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 18. Re: Hit Rating Theorycraft   11-May-2007 11:15:34 PDT
quote reply
Ok, i edited my post and replaced the pvp based example with 2 new images that i took from Ulda my self. Hope i covered you Shvabrus.
70
View All Posts by This User Toggle Ignore / Unignore This User
  • 19. Re: Hit Rating Theorycraft   12-May-2007 03:54:48 PDT
quote reply

Q u o t e:

SPECIAL ATTACKS: Special attacks use a 2 roll system so the crit cap appears to be 100% making Hit less important for them after covering the base miss chance which is 6.5%( no matter if you are a 2hander or dw). Now, let's say you mouseover your crit chance and it says you have 25% crit chance and you also don't have any %hit on you. Your special attacks crit chance against a raid boss (from the back) will be 25% of your hit which makes it to be 25% of 68.9%=17.225% since you don't have any %hit. Increase your hit by 10%. Your new special attacks crit chance will be 25% out of 78.9%=19.725%. Adding more or less hit and doing the maths you'll figure out that every point of hit increases your crit chance of your special attacks by X/100 where X is your crit chance. So basically, HIT INCREASES YOUR CRIT CAP FOR AUTO ATTACKS AND INCREASES YOUR CRIT CHANCE OF SPECIAL ATTACKS.



more proofs please
1 . 2 . 3 . 4 . 5 . 6 . 7 . 8
Forum Nav : Jump To This Forum
Blizzard Entertainment